Promoting physical activity in COPD: Insights from a randomized trial of a web-based intervention and pedometer use

Abstract

Rationale: Low physical activity is highly prevalent among COPD patients and is associated with increased healthcare utilization and mortality and reduced HRQL. The addition of a website to pedometer use is effective at increasing physical activity; however, the timeline of change and impact of environmental factors on efficacy is unknown.

Methods: U.S. Veterans with COPD were randomized (1:1) to receive either (1) a pedometer and website which provided goal-setting, feedback, disease-specific education, and an online community forum or (2) pedometer alone for 3 months. Primary outcome was change in daily step count. Secondary outcomes included 6MWT distance, HRQL, dyspnea, depression, COPD knowledge, exercise self-efficacy, social support, motivation, and confidence to exercise. Generalized linear mixed-effects models evaluated the effect of the pedometer plus website compared to pedometer alone.

Results: Data from 109 subjects (98.5% male, mean age 68.6 ± 8.3 years) were analyzed. At 13 weeks, subjects in the pedometer plus website group had significant increases daily step count from baseline relative to the pedometer alone group (804 ± 356.5 steps per day, p = 0.02). The pedometer plus website group had significant improvements in daily step count from baseline beginning in week 3 which were sustained until week 13. In subgroup analyses, the pedometer plus website attenuated declines in daily step count during the transition from summer to fall. No significant differences in secondary outcomes were noted between groups.

Conclusions: A website added to pedometer use improves daily step counts, sustains walking over 3 months, and attenuates declines in physical activity due to season.

Keywords: COPD; Physical activity; Randomized trial; Rehabilitation; Season.

Conflict of interest statement

Conflicts of Interest Statement

Dr. Moy received an honorarium for consulting from Astra Zeneca. All other authors declare no relevant conflicts of interest.

Published by Elsevier Ltd.

Figures

Figure 1

CONSORT diagram

Figure 2. Change from baseline daily step count by randomization group

Control group = Pedometer alone, Intervention group = Pedometer plus website Values plotted on the y-axis are from least square means solutions for change in daily step count from a generalized linear mixed-effects model for repeated measures, adjusting for FEV1 % predicted and season of intervention week. At week 13, subjects in the pedometer + website group walked an average change in step count from baseline of 804 steps more than subjects in the pedometer-only group (p = 0.02).

Figure 3. Change in daily step count by randomization group among subjects enrolled in summer

Control group = pedometer alone, Intervention group = pedometer plus website Values plotted on the y-axis are from least square means solutions for change in daily step count from a generalized linear mixed-effects model for repeated measures, adjusting for FEV1 % predicted and season of intervention week. Subjects in the pedometer-alone group demonstrated a significant decline in daily step count from baseline during the last three weeks of the study period. Of the 28 subjects enrolled during summer, 23 (82%) ended the study (at week 13) in fall or winter when temperatures in the northeastern United States are cooler.